Randomised controlled trials in oncology closed early for benefit: trends in methodology, results, and interpretation

Profiling clinical cancer research across the Atlantic: a review of research and its characteristics presented at ASCO and ESMO Congresses during the last decade

INTRODUCTION

The comparison of clinical cancer research characteristics across the Atlantic and their evolution over time have not been studied to date.

METHODS

We collected oral presentations on breast, lung and colorectal cancer at ASCO (n=506) and ESMO (n=239) Congresses in years 2000-2010.

RESULTS

EU-originated research constituted 52% of all ASCO presentations while US-research 26.7% of ESMO Congress presentations. Industry sponsorship was reported in 24.8% of ASCO vs. 31.8% of ESMO Congress trials. ASCO-presented trials were larger with longer follow-up periods but were blinded less often. ESMO-presented trials used Event-Free Survival (EFS, 38.1%) and Surrogate (18.4%) primary endpoints and reported positive primary endpoints (65%) more often than ASCO-presented trials. Interim analysis resulted in discontinuation of a trial more often at ASCO Congress (8.3% vs. 3.2%). ASCO Congress-presented research was more often published (69.2% vs. 59.8% at ESMO) at higher impact factor journals. Strong trends over the decade were seen for more frequent industry sponsorship, blinded design, larger sample size, early interim discontinuation, use of EFS endpoints and biomarker evaluation.

CONCLUSIONS

Cancer clinical research is a complex scientific activity with common global but also distinct characteristics at the two sides of the Atlantic.